本文選題：大幅專業(yè)掃描儀　切入點(diǎn)：圖像分割　出處：《西安電子科技大學(xué)》2013年碩士論文　論文類型：學(xué)位論文

【摘要】：隨著大幅專業(yè)掃描儀在各個(gè)領(lǐng)域日益普遍的應(yīng)用，用戶對(duì)掃描精度、圖像質(zhì)量和工作效率要求越來越高，促進(jìn)了大幅專業(yè)掃描儀的不斷創(chuàng)新和研發(fā)。 本文所述的大幅專業(yè)掃描儀采用模塊化的思想將整個(gè)系統(tǒng)分為上層驅(qū)動(dòng)軟件、中間層控制系統(tǒng)和底層圖像數(shù)據(jù)采集系統(tǒng)。本文主要對(duì)掃描儀相關(guān)圖像處理功能進(jìn)行了設(shè)計(jì)和實(shí)現(xiàn)，，位于上層驅(qū)動(dòng)軟件的圖像處理模塊。首先對(duì)圖像分割的相關(guān)理論進(jìn)行了研究和分析；接著在明確掃描儀對(duì)圖像分割功能的具體需求之后，分別設(shè)計(jì)了基于閾值和基于邊緣的圖像分割算法；最后對(duì)兩種圖像分割算法進(jìn)行了實(shí)現(xiàn)，并將圖像分割功能整合到掃描儀上層驅(qū)動(dòng)軟件中。兩種算法都采用由外向內(nèi)逐行（或逐列）搜索分割邊界的思想，其中閾值分割算法利用每行（或每列）的灰度均值進(jìn)行邊界的判定；而邊緣分割算法則基于改進(jìn)的Sobel算子進(jìn)行邊緣點(diǎn)的檢測(cè)來確定邊界位置。本文還對(duì)TIFF和RAW兩種圖像格式進(jìn)行了研究，并在大幅專業(yè)掃描儀中實(shí)現(xiàn)了這兩種格式文件的生成。 經(jīng)過聯(lián)機(jī)調(diào)試，本文設(shè)計(jì)的兩種圖像分割算法都在各自應(yīng)用場景下達(dá)到了預(yù)期的分割效果，有效地優(yōu)化了掃描儀的處理流程，提高了工作效率。掃描儀對(duì)TIFF和RAW兩種圖像格式也得到了很好的支持，擴(kuò)展了大幅專業(yè)掃描儀的應(yīng)用范圍。
[Abstract]:With the increasingly widespread application of large scale professional scanners in various fields, users require more and more high scanning accuracy, image quality and work efficiency, which promote the innovation and development of large scale professional scanners. The large scale professional scanner described in this paper adopts the idea of modularization to divide the whole system into upper driving software. The middle layer control system and the bottom image data acquisition system are designed and implemented in this paper. The image processing module is located in the upper driving software. Firstly, the related theory of image segmentation is studied and analyzed, and then the specific requirements of the scanner for image segmentation are clarified. Image segmentation algorithms based on threshold and edge are designed, and finally, two image segmentation algorithms are implemented. The image segmentation function is integrated into the upper driving software of scanner. Both algorithms adopt the idea of searching the segmentation boundary from outside to inside line by line (or column by column). The threshold segmentation algorithm uses the gray mean of each row (or column) to judge the boundary. The edge segmentation algorithm is based on the improved Sobel operator to detect edge points to determine the edge position. In this paper, two image formats, TIFF and RAW, are studied, and the two formats are generated in a large scale professional scanner. After on-line debugging, the two image segmentation algorithms designed in this paper have achieved the expected segmentation effect in their respective application scenes, and have effectively optimized the processing flow of the scanner. The scanner also has good support for TIFF and RAW image formats and extends the application range of large scale professional scanners.
【學(xué)位授予單位】：西安電子科技大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2013
【分類號(hào)】：TP391.41;TP334.22

【參考文獻(xiàn)】

相關(guān)期刊論文 前4條

1 張新峰,沈蘭蓀;圖像分割技術(shù)研究[J];電路與系統(tǒng)學(xué)報(bào);2004年02期

2 魏弘博,呂振肅,蔣田仔,劉新艷;圖像分割技術(shù)縱覽[J];甘肅科學(xué)學(xué)報(bào);2004年02期

3 袁春蘭;熊宗龍;周雪花;彭小輝;;基于Sobel算子的圖像邊緣檢測(cè)研究[J];激光與紅外;2009年01期

4 趙榮椿,遲耀斌,朱重光;圖像分割技術(shù)進(jìn)展[J];中國體視學(xué)與圖像分析;1998年02期

相關(guān)碩士學(xué)位論文 前5條

1 張朋昌;基于GL847的大幅面圖像數(shù)據(jù)采集系統(tǒng)底層驅(qū)動(dòng)軟件研究與實(shí)現(xiàn)[D];西安電子科技大學(xué);2010年

2 侯武蓉;多CCD掃描儀圖像處理模塊的研究與實(shí)現(xiàn)[D];西安電子科技大學(xué);2010年

3 彭麗;基于邊緣信息的閾值圖像分割[D];中南大學(xué);2009年

4 張素瓊;基于偏微分方程的醫(yī)學(xué)圖像去噪和分割技術(shù)研究[D];浙江工業(yè)大學(xué);2009年

5 曾恒;大幅面多CCD掃描儀圖像處理模塊的設(shè)計(jì)與實(shí)現(xiàn)[D];西安電子科技大學(xué);2012年

本文編號(hào)：1629288